Hi -

I just want to make sure I have the limits correctly calculated for the self test....

1. Are the values in the tables 14 thru 17 in the data sheet limits for the final

output of the self-test-- Xst ?? Ex. Xst = Xst_ON - Xst_OFF

So the Xst difference should be compared to the numbers from the table? Right?

2. For +/- 12g operation at VDD = 3VDC we have to scale the limits from

Table 17 since they are shown for 10 bit operation, 2.5v, and we are at 13 bit, 3volts operation.

So, Table 17 shows limits for 12g at 10 bit, 2.5v : X = 8 to 90

But we want to convert to 13 bits and 3.0 volts so for example

Lower limit = 8 bits @ 10 bit reso. = 0.0234375g/bit * 8 = 0.1875 g.

Now scale by 1.77 for 3 volts supply then = 0.331875g

Now divide by how much g in 1 bit at 13 bit (signed) full resolution:

For +/-12g range --> 12g / 4096 = .0029297g/bit

This gives a new X axis lower limit value of 0.33187 / .0029297 = 113.277 (truncate to 113) bits

and an upper limit of 1274 (for the original 90 value)

Is this the correct methodology?

Thanks!

Hi,

1. You are correct that the stated limits for self test are the result of the difference of the 2 measurements taken.

2. You should use Table 1 as the correct limit for self test response per axis. The high resolution mode is better for this to provide a more accurate measurement:

So min X Axis self test response should be 0.2 g/ (2.9mg/LSB) ~69 LSB.

Similarly, Y axis this should be a shift of at least - 69 LSB or more negative.

Z Axis should shift 0.3 g/(2.9 mg/LSB) ~ 103 LSB.

As a sanity check (roughly!) the 13 bit limits should scale by about 8 (2^3) from the 10 bit limits shown in Table 17. It is better to do the self test with the 13 bit mode, exactly the way you are doing it because of the improved resolution.